Higher rates, from pensions to apartments · Morning Gist
Sunday, October 255 stories · 15 min
Higher rates, from pensions to apartments
Energy pushed September inflation up, but the core trend stayed near 3–4%, and that is showing up in pension payouts, landlords' loans and heirs' tax bills.
Picked from 33 new articles across 16 sources · arrow keys for other days
Today in 30 seconds
30s left
Energy pushed September consumer prices up 8.1% annualized. Without volatile items, inflation stays near 3–4%, which points to rates staying higher for longer. story 1 ↓
Heirs of many inherited IRAs must take withdrawals every year without reminders. In one example, a widow who skipped four years lost about $47,700. story 2 ↓
Rising interest rates cut one pension's lump-sum offer by about 8%. The plan resets its rates only once a year, so the payout jumps between updates. story 3 ↓
The apartment survey's debt financing index fell to 10, and 17% of affected landlords say some properties may face foreclosure as refinancing gets harder. story 4 ↓
A researcher estimates 720,000 people work in cybercrime but only a few thousand are skilled hackers, so hiring limits how much crime gets done. story 5 ↓
A paper built from 20,000+ calendar entries shows how Bernanke, Yellen and Powell split their time among markets, politics and policy work.
1/5The Overshoot· Matthew C. Klein· Oct 22
September inflation jumped on energy, but the underlying trend held near 3–4%
September's U.S. inflation figures looked alarming because of energy prices. Matthew Klein argues the steadier underlying rate, about 3–4% a year, is the real story: higher than before the pandemic, and not falling back toward 2%.
Why it matters · If underlying inflation has settled near 3–4%, savers, borrowers and bond investors should plan for interest rates to stay higher for longer, not for a return to 2%.
Know this first
Context from the editor, not the article · 28 sec
1 · Headline vs. underlying inflation
Headline inflation counts every price. Economists also track "core" or underlying measures that leave out volatile items such as food and energy, which can swing a lot from month to month without reflecting the broader trend.
2 · Annualizing monthly numbers
A one-month price change is often restated as the yearly rate it would add up to if it repeated every month. That makes it comparable with the Federal Reserve's 2% annual target, but it also magnifies one-off spikes.
3 · The Fed's 2% goal
The Federal Reserve, the U.S. central bank, aims for 2% annual inflation. It mainly steers by raising or cutting its short-term interest rate target, which feeds through to borrowing costs across the economy.
2/5HumbleDollar· John Urban· Oct 24
Inherited IRAs often need yearly withdrawals, and no one sends a reminder
John Urban warns that many heirs must take yearly withdrawals from inherited IRAs, and banks aren't required to remind them. Missing those withdrawals brings a 25% penalty plus a delayed Medicare premium surcharge.
Why it matters · Trillions of dollars in retirement accounts will pass to heirs. One date you can look up today decides whether an heir owes withdrawals every year or can wait.
Know this first
Context from the editor, not the article · 32 sec
1 · Required minimum distributions
Traditional IRAs grow tax-deferred, so the IRS forces withdrawals, called required minimum distributions (RMDs), starting at a set age. Each year's amount is the prior year-end balance divided by an IRS life-expectancy factor. Withdrawals are taxed as ordinary income.
2 · The SECURE Acts
The 2019 SECURE Act ended the lifetime 'stretch' for most non-spouse heirs. They now must empty an inherited IRA within 10 years. SECURE 2.0, passed in 2022, raised the age when owners must start withdrawals and cut the penalty for missed withdrawals.
3 · Rule confusion, then enforcement
3/5HumbleDollar· thehalperin· Oct 22
Rising interest rates cut a pension lump-sum offer by about 8%
A former portfolio manager found the lump-sum value of his old company pension had fallen about 8% in a year. The cause: rates rose more than one percentage point, so his future payments are worth less today.
Why it matters · If you plan to cash out a pension, the amount you'll get moves like a bond price, so a fixed number in your retirement plan could be well off.
Know this first
Context from the editor, not the article · 29 sec
1 · Defined benefit pensions
A defined benefit (DB) plan promises a set monthly income in retirement, usually based on years of service and pay. Many plans also let you swap that income stream for a one-time cash payment, the lump sum.
2 · Present value
Money paid in the future is worth less than money today, because today's money can earn interest. Present value turns future payments into one amount today by discounting them at an interest rate. Higher rates mean a lower present value.
3 · Bond prices and rates
When market interest rates rise, existing bonds lose value, because their fixed payments now look less attractive. The longer the payments stretch out, the bigger the price change. Pension lump sums work the same way.
4/5Calculated Risk· CalculatedRisk by Bill McBride· Oct 22
Apartment market loosens as rising rates squeeze landlords' refinancing
An October survey of apartment executives by the National Multifamily Housing Council found looser rental markets, few property sales and scarce financing. Higher rates leave many landlords unable to refinance loans taken out in 2020–21.
Why it matters · The survey has been a good early signal for rents and vacancies, and it shows how higher long-term rates feed through to property owners, lenders and future housing supply.
Know this first
Context from the editor, not the article · 27 sec
1 · What NMHC's survey measures
The National Multifamily Housing Council, an apartment-industry trade group, asks executives each quarter whether conditions are better, worse or unchanged than three months earlier. Each answer set becomes an index where 50 means no change.
2 · Why rates hit landlords
Apartment buildings are mostly bought with borrowed money. Long-term loan rates follow the 10-year Treasury yield, so when that yield rises, borrowing gets more expensive and buyers will only pay lower prices for buildings.
3 · The 2020–21 loan wall
Commercial property loans usually run five to ten years and end with a large balance still owed, which has to be refinanced. Loans taken out at record-low pandemic-era rates are now reaching that point.
5/5Marginal Revolution· Tyler Cowen· Oct 24
Cybercrime is limited by who it can hire, a security researcher argues
Tyler Cowen quotes Joshua Saxe on why cyber losses aren't bigger. Saxe's answer is the labor market: criminal gangs struggle to hire skilled people, so most of their workforce is low-wage scam labor, not elite hackers.
Why it matters · It applies basic labor economics, wages and opportunity cost, to a risk that touches your bank account and that insurers and companies price every year.
Know this first
Context from the editor, not the article · 19 sec
1 · Economics of crime
Economists since Gary Becker have treated crime as a career choice. People weigh what crime pays, and the risk of getting caught, against what they could earn legally. Better legal options mean fewer recruits.
2 · Ransomware vs. scams
Ransomware means breaking into computer systems, encrypting the data and demanding payment to unlock it, which takes real technical skill. Scam and fraud call centers mostly need people willing to phone or message victims, and very little technical skill.
AI safety
Field focused on preventing harmful uses and behaviors of artificial intelligence systems.
The U.S. 1, the government's main measure of prices paid by households, rose 0.65% from August to September after 2. That works out to an 8.1% annual rate. The 3 for final demand, which tracks prices businesses charge for finished goods and services, rose 1.34% in the month, or 17.4% 4. That is its fastest pace since March 2022. Retail and restaurant spending rose 0.77% in September, following a 1.29% jump in August.
Mostly energy
Matthew Klein of The Overshoot argues the scary headline figures are mostly about energy. Leave energy out and CPI rose at about a 4% annual rate, the same as in August. Also leave out other volatile or hard-to-measure categories and the rate falls to roughly 3–3.5%. On the producer side, consumer prices excluding food, energy and "5" such as trucking and flights rose at a 4.5% annual rate.
These underlying rates are far below the headlines. Klein's point is that they are still clearly above pre-pandemic inflation, and they have been "remarkably consistent" for several years.
No easy road back to 2%
This supports the view Klein has held for years. The pandemic and the policy response to it reset underlying inflation a bit higher. That means there will be no "6" disinflation, meaning a painless drift back to the Federal Reserve's 2% target without the economy slowing.
Klein agrees the current administration's tariffs and the war with Iran have pushed prices higher. But he calls it misleading to claim inflation would have fallen to 2% without them. Services prices have been rising faster than before the pandemic for years. Goods prices were never falling fast enough to offset that, even before the tariffs, the data-center boom and the war.
Markets moved on the data. 7's futures-based estimate of the chance that the Fed raises rates by half a percentage point by year-end went from 13.3% on October 7 to 28.5% after the CPI release on October 14. It has since fallen to 5.6%, on worries about how much some data-center operators are borrowing.
What to watch
Some good news: the underlying trend doesn't look to be getting much worse, and the sharpest jumps in manufacturing-input prices may have peaked.
Wages: pay growth for the typical worker has slowed a little. Will that slowdown feed through to spending and prices, or reverse now that hiring is picking up again?
Supply shocks: given how retailers have managed their inventories, are prices now more or less sensitive to new supply disruptions?
The available text stops before Klein makes the "consumers are fine" argument in the title, so his case for it isn't covered here.
September's U.S. inflation figures looked bad on the surface. The 1, the Bureau of Labor Statistics' measure of prices paid by households, rose 0.65% from August after 2. Spending at retailers, bars and restaurants rose 0.77% in September, after a 1.29% jump in August. The 3 for final demand, which tracks prices U.S. businesses charge for finished goods and services, rose 1.34% in a single month. That was its fastest rise since March 2022.
Matthew Klein of The Overshoot reads these numbers as confirming his long-held view rather than signalling a new problem. In his telling, inflation is not getting much worse, and it is not getting better either.
Why monthly numbers look so dramatic
Klein reports most of these figures "4." That means the one-month change is compounded as if it happened every month for a year. Here is the arithmetic for CPI: 1.0065 raised to the 12th power is about 1.081, so a 0.65% monthly rise becomes 8.1% a year. The same method turns PPI's 1.34% into 17.4% and August's 1.29% spending gain into 16.6%.
Annualizing makes a month's data comparable with the Federal Reserve's 2% annual target. It also magnifies anything temporary, which is why separating noise from trend matters so much.
Taking out the noise
Klein says much of September's jump came from energy goods and related items. He peels the data back in layers:
"Distributive services" are things like trucking and airline flights. Their producer prices swing with fuel costs and margins. Klein also points out that in August some analysts blamed wireless-plan prices for the strong reading. In September those prices fell, so that explanation doesn't work this time.
His conclusion has two parts. The underlying rates are far below the alarming headlines. But they are clearly faster than inflation just before the pandemic, and they have been "remarkably consistent" for several years.
The argument: a reset, not a blip
Klein's thesis is that the pandemic and the policy response to it reset underlying inflation a little higher. If so, there will be no "6" disinflation. That term means inflation falling back to 2% without any pain, such as weaker hiring or tighter policy.
The mechanism he describes is a mismatch between services and goods. Underlying services prices have been rising persistently faster than before the pandemic. Goods prices were never falling fast enough to offset them. That was true, he stresses, before the tariffs, the boom in data-center construction, and the war with Iran.
This leads to his political point, which he states carefully. Klein says it is "understandable," even "only fair," to blame the current administration. The 2024 election was won partly on what he calls misplaced anger about inflation and "ludicrous claims" that new leadership would make prices fall. He also accepts that the tariffs and the war with Iran have made prices higher than they would otherwise be. But he calls it misleading to say inflation would have eased to 2% on its own without those policies. In his view the trend was already stuck above target.
How markets reacted
Klein follows the 7's futures-implied probabilities, which back out market expectations for Fed policy from interest-rate futures prices. The market's estimate of the chance that the Fed raises its short-term rate target by half a percentage point by year-end moved like this:
13.3% on October 7
18.7% on October 13
28.5% at the close on October 14, after the CPI release
5.6% now, after concerns about how much credit some data-center operators need
That last move matters. A large borrower running into trouble can tighten credit conditions and slow the economy. So markets can quickly stop pricing in a rate hike even when the inflation data alone would argue for one.
Some good news, and open questions
Klein sees some relief. The data don't suggest the underlying trends are getting much worse. Some of the sharpest price spikes, in the PPI categories "components for manufacturing" and "supplies to manufacturing industries," may have peaked.
He raises two questions without answering them in the available text:
Wages. Inflation has held steady or crept up slightly, while the typical worker's wage growth has slowed modestly. If that slowdown continues, it could eventually show up as weaker spending and slower price increases. But the job market has recently picked up again, which could push wage growth back up.
Inventories and supply shocks. Given what is now known about how retailers managed their stock in recent years, are prices more or less sensitive to new supply disruptions? Holding more stock can cushion a shock. Running lean can make prices jump faster.
What this account can't tell you
The text provided cuts off partway through the post. Klein's argument for the second half of his title, that "consumers are fine," isn't included, beyond the strong August and September spending figures. The cut-off text also doesn't explain the wage figures behind his claim that pay growth has slowed, or show how he builds his "volatile, idiosyncratic" exclusion measure. The 3–3.5% figure is his own construction, not an official series.
For a reader planning their finances, the practical takeaway is Klein's main claim. Underlying inflation of roughly 3–4% has proven stable, and he sees no sign of it drifting down to 2% on its own. If he is right, interest rates are more likely to stay high than to fall quickly.
Remember
Annualizing one month's data magnifies spikes: September's 0.65% CPI rise becomes an 8.1% annual rate.
Services prices have stayed high since the pandemic, and goods prices haven't fallen enough to offset them.
Market odds of a Fed rate hike can swing sharply on credit worries, not only on inflation data.
After 2019, the IRS ruled that some heirs on the 10-year clock also owe yearly withdrawals. It waived penalties on missed ones from 2021 to 2024. Final rules require these yearly withdrawals from 2025 on, so the risk is now real.
Image: HumbleDollar
No reminder for heirs
Every January, IRA 1 (the banks and brokerages holding the accounts) tell owners how much they must withdraw that year. John Urban argues on HumbleDollar that this letter, not diligence, is why most people never miss a 2, or RMD, the yearly withdrawal the IRS forces from tax-deferred accounts.
Inherited IRAs don't come with that letter. The IRS notice behind the custodian reporting requirement exempts IRAs of deceased owners, so the heir often hears nothing.
What the 10-year rule leaves out
Under the 3, the 2019 law governing inherited accounts, most non-spouse heirs must empty an inherited IRA within 10 years. That rule doesn't say whether something must come out each year. The answer depends on whether the person who died had reached their own RMD age: 72 if born before 1951, 73 if born 1951–1959, 75 if born 1960 or later.
If an adult child's parent died before that age, the heir can wait until year 10. If the parent died on or after it, the heir owes an RMD every year and must still empty the account by year 10. Urban says this is the case people get wrong.
A shortfall carries a 25% 4, a penalty on top of income tax, or 10% if corrected promptly. Money left in the account at the end of year 10 also counts as a shortfall.
A worked example
A 76-year-old widow inherits her 78-year-old husband's $480,000 IRA and leaves it titled as inherited. She has $60,000 a year from Social Security and a pension, takes nothing for four years, and the account grows 5% a year to $555,660.
Missed withdrawals: $160,631.
Excise tax: $40,158 at 25%, or $16,063 at 10%.
Extra income tax from bunching four withdrawals into one year: only $1,216 ($38,783 versus $37,566 if taken on schedule). Urban says the feared bracket jump is mostly a myth.
Medicare surcharge: $6,355. Higher earners pay extra for Medicare Part B and Part D, based on income from two years earlier, so this bill shows up long after the fix.
Her avoidable cost is about $47,700. The fix also means amended returns and a 5, the IRS penalty form, for every missed year.
What to do
If you've inherited, find two facts: the year the owner died and whether they had reached RMD age by then. If you're married and both spouses are alive, Urban suggests writing down the year each of you reaches RMD age, with a note on why it matters, and filing it with your will and beneficiary designations.
The article doesn't cover IRS waivers for reasonable cause or the spousal options that would have changed the widow's timeline. Urban runs a retirement tax-planning app.
The letter that never comes
Every January, IRA 1 (the banks and brokerages that hold the accounts) write to owners who must withdraw money that year. The letter gives the amount or offers to work it out. John Urban, writing on HumbleDollar, argues that this letter is the main reason most people never miss a 2, or RMD, the withdrawal the IRS forces from tax-deferred accounts. People don't remember the rule. Someone reminds them.
That safety net is missing for inherited IRAs. The IRS notice that created the custodian reporting requirement says reporting isn't required for IRAs of deceased owners, and an inherited IRA is exactly that. The custodian owes the heir no letter, and often none arrives. Urban's point is that when nobody says anything needs doing, people assume nothing does.
The question the 10-year rule doesn't answer
Most heirs have heard of the 10-year rule from the 3, the 2019 law that changed how inherited retirement accounts are paid out. It says when the account has to be empty: by the end of the tenth year after the death. It doesn't say whether money has to come out this year, and that question is the one that costs money.
The answer depends on one fact about the person who died: had they reached their own RMD starting age? That age is 72 for people born before 1951, 73 for those born 1951–1959, and 75 for those born in 1960 or later.
In the common case of an adult child inheriting a traditional IRA:
Parent died before that age: nothing has to come out until the account is emptied by the end of year 10.
Parent died on or after it: an RMD is due every year, and the account still has to be empty by the end of year 10.
Urban says the second case trips people up. They remember the 10-year rule as permission to leave the money alone, and for many heirs it isn't. Other heirs have their own timing: disabled or chronically ill beneficiaries, minor children, and anyone less than 10 years younger than the person who died. Inherited Roth IRAs work differently again, and spouses have options nobody else has. The rules are in the SECURE Act and its 2022 follow-up, 4.
The penalty
A shortfall triggers an 5, a penalty tax on top of income tax, equal to 25% of the amount that should have come out. It drops to 10% if corrected promptly. You still owe ordinary income tax when the money finally comes out. There are two ways to trigger it: taking too little in a given year, or reaching the end of year 10 with money still in the account. Whatever is left on that date also counts as a shortfall.
A worked example: four quiet years
Urban builds a case. A widow turns 76 the year after her husband dies at 78, well past his RMD age. She has $36,000 of Social Security, a $24,000 pension, and his $480,000 IRA, which she leaves titled as an inherited account rather than taking one of the spousal options. She doesn't need the money, no reminder comes, and she withdraws nothing for four years while the account grows 5% a year.
Because he had passed his RMD age and the account stayed titled as inherited, she owed a withdrawal every year:
Her age
Required withdrawal
76
$34,043
77
$37,895
78
$42,000
79
$46,694
By the end, the untouched balance is $555,660 and she has missed $160,631 in required withdrawals. At 25%, the excise tax is $40,158. Had she qualified for the prompt-correction rate of 10%, it would have been $16,063.
The bracket worry is mostly a myth
When she corrects the mistake, all four withdrawals come out in one tax year. Urban expected this bunching to be the hidden disaster, since it pushes income into higher brackets. It barely matters. Taken on schedule, the four withdrawals would have cost $37,566 in federal income tax. Bunched into one year, they cost $38,783, a difference of $1,216. He tells readers to check claims about bracket damage against their own numbers.
The Medicare surcharge
The bigger surprise is Medicare. Higher earners pay a surcharge on Part B (doctor visits) and Part D (prescription drugs) premiums. It's formally called 6, the income-related monthly adjustment amount, and it's based on income from two years earlier. Taken on time, her withdrawals trigger no surcharge in any year. Bunched into one return, they trigger $6,355. That bill arrives two years after the fix, attached to premiums that don't look like a tax.
Her avoidable cost totals about $47,700: $40,158 in excise tax, $1,216 in extra income tax, and $6,355 in Medicare surcharges. The $37,566 of ordinary income tax was always going to be owed. In Urban's telling, the first letter she gets comes from the IRS years later, and fixing it means amended returns and a 7 (the IRS form for reporting retirement-account penalties) for every year.
What prevents it
Urban argues that vigilance won't help, because nobody is vigilant about a rule they've never heard of. For people who have already inherited, two facts settle most cases: the year the owner died, and whether they had reached their RMD age by then. The custodian can tell you how the account is titled. Only you can find that second date.
For married couples who are both alive, he recommends working out the year each spouse reaches RMD age now. For whoever dies first, that year decides whether the survivor owes a withdrawal the year after the death. Write both years down with a note explaining why they matter, and keep it with the will, the trust and the beneficiary designations, where a survivor will actually find it.
Caveats
The example is one constructed case. It doesn't show how the RMD factors or the tax figures were calculated, and it doesn't say whether the widow could have gotten the 10% rate or a penalty waiver. The IRS can waive the penalty for reasonable cause, which the article doesn't discuss. The article also doesn't walk through the spousal options that would have changed her timeline. Urban founded RetireSmartIRA, a retirement tax-planning app, so he works in this field. That's worth knowing, though the piece doesn't sell anything.
Remember
Whether yearly withdrawals are owed depends on whether the person who died had reached their own RMD age.
Bunching missed withdrawals into one year barely raised income tax here; the penalty and the Medicare surcharge did the damage.
Married couples can record each spouse's RMD-age year now so the survivor knows which rules apply.
Matt Halperin, a 1 and former portfolio manager, has a small pension from a company he left decades ago. It's a 2 plan, which pays a fixed monthly income based on his years of service and pay. At 65 he can take those payments or cash out in one lump sum. On his yearly check of the account, he saw the lump-sum offer had fallen about 8%.
Why it fell
The plan sets the lump sum as the 3 of his expected lifetime payments. That means it discounts each future payment back to today at an interest rate tied to the bond market. That rate rose by more than 100 4 (one percentage point) over the year. Halperin adds that an updated 5, the plan's estimate of how long people live, may also have played a part.
An illustrative example: $1,000 a month for 20 years is worth about $151,500 today at a 5% rate, but about $139,600 at 6%. That's roughly 8% less. Payments that start further in the future swing even more.
Doesn't the higher rate make up for it?
You could invest the smaller lump sum at today's higher yields. In the example, $139,600 invested at 6% gives back about $1,000 a month, the same as the pension. Halperin says the offset is "likely but not necessarily." It depends on three things:
whether you can actually earn the rate the plan used (it may assume riskier corporate bonds),
what investments are open to you,
and when the cash reaches your account.
General background the article doesn't give: US rules base minimum lump sums on corporate bond rates and IRS mortality tables. The math also assumes average lifespan, so someone who lives long gives up guaranteed income.
Rates reset once a year
His plan updates its rates only once a year. If market rates fall before he cashes out, the next reset will raise the lump sum, and he'd miss that gain by taking today's figure. If rates rise, cashing out now looks better in hindsight. Halperin's advice: don't treat a planned lump sum as a fixed number. Treat it as "a bond whose value will fluctuate with interest rates."
Don't lose track of old plans
Halperin runs Act2 Financial, an app that helps seniors avoid fraud. He suggests checking old pension and 401(k) accounts once a year. Confirm your contact details, the correct website and your login, and watch for account takeovers. Also make sure your spouse or your written instructions for survivors explain how to claim any death benefit. An old plan that sends one letter a year is easy to forget.
What happened
Matt Halperin, a 1 who spent 30 years as a portfolio and risk manager, worked six years early in his career at a company with a 2 pension. A DB plan pays a fixed monthly income based on a formula, here years of service and his pay at the time. He stayed just long enough to vest, meaning he earned a permanent right to the benefit. At 65 he can take either the monthly payments or a one-time lump sum.
He logs in once a year to check the account. On his latest visit, writing on HumbleDollar, he found the lump-sum figure had dropped by about 8%. His monthly benefit hadn't changed. What changed was the price the plan puts on that benefit.
How a lump sum is calculated
Halperin explains that the plan sets the lump sum as the 3 of his monthly payments over his expected lifetime. Present value is today's worth of future cash. You get it by discounting each future payment at an interest rate. Two inputs drive the result:
The discount rate. In his plan it tracks the bond market, and it rose by more than 100 4 (one percentage point; a basis point is 0.01%) over the past year.
Expected lifespan. This comes from an 5, a statistical estimate of how long people live. Halperin says changes to that table may also have played a part.
Some general background the article doesn't give: for US private-sector plans, federal rules (Internal Revenue Code section 417(e)) set a floor on lump sums. That floor is calculated with IRS-published corporate bond rates and IRS mortality tables. Many plans fix those rates for a whole plan year, which fits Halperin's note that his plan updates them only once a year.
A worked example (illustrative numbers)
Say a pension pays $1,000 a month for 20 years.
Discounted at 5%, those 240 payments are worth about $151,500 today.
Discounted at 6%, they are worth about $139,600.
A one-point rise in rates cuts the value by roughly 8%, close to what Halperin saw. Bond investors would call this an asset with a 6 of about 8. Duration measures price sensitivity: roughly, the percentage drop in value for each one-point rise in rates. If payments start years from now, as they do for someone not yet 65, the cash is further away and the value swings even more.
Doesn't a higher rate make up for it?
Halperin raises the obvious objection himself. If rates are higher, he could invest the smaller lump sum in bonds at the higher yield. Doesn't that cancel out the loss?
His answer is "likely but not necessarily." The example shows why the logic mostly holds. Invest $139,600 at 6% and draw it down over 20 years, and you get back almost exactly $1,000 a month. That's no accident: the discount rate that shrank the lump sum is the same rate that rebuilds the income. The offset is only exact, though, if you can really earn the rate the plan used. Halperin lists what decides it:
The rate used in the formula. The plan's rate may not match anything you can actually buy.
The investments open to you. Corporate bonds carry default risk. Treasuries are safer but usually pay less.
Timing. What matters is when the money lands in your brokerage account. Rates can move between the plan's annual reset and the day you invest.
One more gap, from general knowledge rather than the article: the lump-sum math uses average life expectancy. Someone who lives well past it would have done better keeping the lifetime payments. A lump sum can run out; the monthly pension can't.
The once-a-year reset
Because his plan updates its rates only annually, the lump sum is effectively locked until the next reset. If market rates fall before he cashes out, the next reset will raise the lump sum, so cashing out at today's figure means missing that gain. If rates rise, the next figure will be lower, so cashing out now looks better in hindsight. For anyone near the decision, knowing when the reset happens and which month's rates it uses matters.
Halperin's main advice is about planning. Don't put the lump sum into your retirement plan as a fixed number. Treat it as a variable amount, in his words "a bond whose value will fluctuate with interest rates." Practically, run your plan at several possible lump-sum values, not one.
The housekeeping point
Halperin now runs Act2 Financial, an app that helps seniors avoid financial fraud, and he adds some practical advice. Check old DB plans, and old 7 plans like 401(k)s, once a year. Confirm your contact details, the correct website and your login. Watch for signs a fraudster has taken over the account. Make sure your spouse or your written "final directions" (instructions for survivors) say how to claim any death benefit. His old plan sends about one piece of mail a year, which is why such benefits are easy to forget.
What the article leaves open
The piece is a short personal note, not an analysis. It doesn't give the plan's discount rate, its mortality table, or how much of the 8% came from rates versus table changes. It also doesn't compare taking the lump sum with keeping the annuity in his own case, beyond saying the answer depends on details. The useful part is the framework: a pension lump sum behaves like a long bond, and you should plan for it that way.
Remember
A pension lump sum acts like a long-term bond: as a rough guide, each one-point rate rise can cut it by about 8%.
Higher yields partly offset a smaller lump sum, but only if you can earn the rate the plan assumed.
Find out when your plan resets its rates; that date can change how much you get.
NMHC's quarterly Apartment Market Tightness Index over time, now at 47; below 50 signals loosening.
The 1, the apartment industry's trade group, released its October quarterly survey. It shows rental markets loosening and financing drying up. Bill McBride of Calculated Risk highlights it because the survey has historically given early warning of where rents and vacancies are heading.
How the indexes work
Executives say whether conditions improved, worsened or stayed the same compared with three months ago. Each index is centred on 50: a reading above 50 means more people saw improvement, and a reading below 50 means more saw deterioration.
The readings
Market Tightness: 47, down from 57 in July. 26% said conditions were looser and 21% said tighter. 53% saw no change, so this is a mild loosening.
Sales Volume: 22. 61% saw fewer deals and 4% saw more.
Equity Financing: 17. 66% said investor money was harder to raise.
Debt Financing: 10. 81% said it was a worse time to borrow.
The cause is rates. The 10-year Treasury yield averaged 5.23% so far in October. That is 63 2 (0.63 percentage points) above July and 117 above a year earlier. NMHC chief economist Chris Bruen says higher inflation drove the jump in long-term rates. Buyers now demand lower prices "to make their deals pencil," so fewer sales get done.
The refinancing problem
Many owners borrowed at very low rates in 2020 and 2021 and now have to refinance loans that are coming due. Lenders cap new loans in two ways. The 3 requires a building's income to exceed its loan payments by a margin. The 4 limits the loan to a share of the property's value. Higher rates raise the payments and push values down, so the largest new loan a lender will allow can come in below the balance still owed.
78% of respondents have loans maturing. Among them:
61% say those ratios stop them refinancing the full amount.
45% will put in more of their own cash.
39% are getting extensions or modifications from lenders.
16% will use 5 or 6, costlier funding that sits between the bank loan and the owner's stake.
17% say some properties may face foreclosure.
What it signals
McBride reads the tightness index as a sign that vacancies and asking rents will stay under pressure. Monthly rent data show tentative signs that asking rents are near a bottom, though. He says this survey and the 7, an early gauge of building design work, both point to further weakness in new apartment construction.
The 1, the trade group for large apartment owners, developers and lenders, published its October Quarterly Survey of Apartment Market Conditions. Every index it tracks came in below 50, and the financing measures fell close to their floor. Bill McBride of Calculated Risk follows the survey because its main gauge, the Market Tightness Index, has historically turned before rents and vacancy rates do.
How a 2 works
Each index is a diffusion index. It doesn't measure rents or prices directly. It summarises how many respondents say things got better or worse over the past three months. A reading of 50 is the breakeven point, and the published numbers match the standard formula: take the share saying "better," subtract the share saying "worse," halve the result and add 50.
Worked check using the October figures:
Market Tightness: 21% tighter minus 26% looser = −5. Half of that is −2.5, plus 50 gives 47.5, published as 47.
Debt Financing: 2% better minus 81% worse = −79. Half is −39.5, plus 50 gives 10.5, published as 10.
So the index tells you about breadth (how many firms feel a change), not how big the change is. That matters for reading the tightness result. A 47 with 53% reporting "unchanged" describes a market that is softening slightly, not collapsing.
The October readings
Market Tightness: 47, down from 57 in July. Below 50 means looser conditions: higher vacancies and less pricing power for landlords.
Sales Volume: 22. 61% saw fewer transactions, 4% saw more and 35% saw no change.
Equity Financing: 17. 66% said money from investors was less available, 1% said more available and 28% said unchanged.
Debt Financing: 10. 81% called it a worse time to borrow, 2% better and 17% unchanged.
The gap between the readings is the story. Day-to-day rental demand is only slightly softer. The capital markets that fund buying, building and refinancing apartments have seized up.
Why: long-term rates
The 10-year Treasury yield averaged 5.23% so far in October. That is 63 3 above July and 117 above a year earlier (a basis point is a hundredth of a percentage point). Commercial mortgage rates are priced off this yield. Equity investors also compare property returns against it, so a higher risk-free return means they want more from a building before they'll buy it.
Chris Bruen, NMHC's chief economist, ties the spike to higher inflation pushing up long-term rates. That raised the cost of both debt and equity. Buyers have responded by demanding lower valuations "to make their deals pencil," meaning to make the numbers work. Sellers often won't accept those prices yet, so deals stall. That is why the sales index sits at 22.
The refinancing squeeze
Commercial property loans typically last five to ten years and end with most of the balance still owed, so owners have to refinance when the loan matures. Many borrowed at extremely low rates in 2020 and 2021. Lenders size a new loan using two limits:
Debt service coverage ratio (DSCR): the property's net operating income (rent minus operating costs) divided by the annual loan payments. Lenders commonly require something like 1.25, meaning income covers payments with a 25% cushion.
Loan-to-value ratio (LTV): the loan as a share of the property's appraised value.
An illustrative example (the numbers are ours, not the survey's). A building earns $600,000 a year in net operating income. In 2021 investors valued it at a 4.5% yield, about $13.3 million, and the owner borrowed $9.3 million (70% LTV) at 3%, paying roughly $280,000 a year in interest-only payments. Now that loan matures. At 7%, interest alone on $9.3 million is about $650,000, more than the building earns, so the DSCR is below 1. If buyers now want a 6% yield, the building is worth about $10 million, and a 65% LTV cap allows only $6.5 million. The owner is roughly $2.8 million short, even though rents haven't fallen.
The survey asked about exactly this situation. 22% of respondents have few loans maturing and are largely unaffected. Of the other 78%:
61% say DSCR limits (54%) or LTV limits (39%) stop them refinancing the full amount.
45% will inject additional equity, meaning more of their own cash.
One inconsistency in the numbers: Bruen's quote says 53% of respondents can't refinance the full amount. The breakdown's 61% of 78% works out to about 48% of all respondents. The release doesn't explain the difference, which may come from rounding or a different base.
What it means
McBride says the tightness index has been an excellent leading indicator. Its drop suggests vacancy rates and asking rents remain under pressure. He balances this with monthly rent surveys showing "emerging signs" that asking rents are near a bottom. He also pairs the survey with the 6, a monthly measure of demand for architects' design work that tends to lead construction by about a year. Both point to further weakness in multifamily starts, meaning new apartment buildings breaking ground. That would cut future supply, which could eventually help rents firm.
What's left open
The survey reports sentiment, not loan performance. It can't say how many properties will actually be foreclosed, or how much lenders' "for the time being" extensions just postpone the reckoning in the hope that rates fall. It also gives no sample size or breakdown by region or property type. The squeeze is likely uneven, landing hardest on owners who bought at peak prices with high leverage. What happens next depends mostly on where the 10-year yield goes, which neither the survey nor McBride forecasts.
Remember
Rental demand is only slightly softer, but financing has nearly frozen: the debt financing index is at 10.
Owners can be short on refinancing even with steady rents, because higher rates both raise payments and lower property values.
Weak financing and design activity point to fewer new apartment starts, which could tighten supply later.
Tyler Cowen asks why cybersecurity losses aren't worse. Software is full of holes, and attackers are often described as brilliant and well organized. On his blog 1, he points to an answer from security researcher Joshua Saxe, which he calls "very good throughout."
Crime has a hiring problem
Saxe starts from labor economics. Anyone building a criminal gang, he writes, must find people "willing to forgo more stable opportunities in the 2." Skilled people with good legal job offers cost a lot to recruit. People with few options cost little.
So gangs base themselves in middle- and lower-income countries, where wages are low and legal prospects are thin. Saxe's examples: an illegal scam call center in a small Nigerian city, and 3 "kids working in the informal economy in Belarus." His piece includes a chart comparing attackers' wages with local wages, but Cowen's excerpt doesn't reproduce the figures.
The headcount
Saxe's model estimates:
About 720,000 people work in online crime.
About 90% of them work in scam and fraud call centers, some of them 4, meaning forced into the work.
About 76,000 are actual hackers, and the "vast majority" of those work for governments.
Only a few thousand are criminal hackers, mostly in ransomware.
The excerpt doesn't show how these numbers were built, so treat them as rough orders of magnitude.
Why that limits the damage
Sophisticated attacks need scarce skills, and most people with those skills have government jobs or good legal careers. That leaves a small criminal talent pool. The large criminal workforce does low-skill scam work, which hurts victims one at a time but can't scale up advanced hacking.
Saxe argues that security and AI-safety people tend to picture attackers as their "5", an evil mirror image of themselves, "implicit MIT graduates." In reality, he says, cybercrime draws on "a lower-skill demographic."
What's open
The excerpt ends by noting "bottlenecks on the down side too" but doesn't explain them. Two inferences of our own, not Saxe's claims. First, if AI tools let low-skill workers do expert work, this protective bottleneck could weaken. Second, for ordinary people the likeliest threat is a scam call or message, not a clever break-in.
The puzzle
Economist Tyler Cowen, on his blog 1, asks why cybersecurity losses aren't worse. Software is full of holes, a lot of money moves online, and people in AI and security circles often describe attackers as brilliant and well organized. So why isn't the damage even bigger? Cowen points readers to an answer from Joshua Saxe, a security researcher. Cowen calls the piece "very good throughout" and quotes the first part of it.
Saxe's answer: crime has a hiring problem
Saxe starts from labor supply. Anyone building a criminal gang, he writes, "has to find candidates willing to forgo more stable opportunities in the 2." That is the standard economics of crime. A person picks crime only when it pays more than their legal options, after allowing for the risk and instability. People with good degrees and good job offers are expensive to lure away. People with few options are cheap.
So, Saxe argues, criminal organizations base themselves in middle- and lower-income countries. Wages there are lower, and many workers have few prospects in the formal economy. He gives two pictures. For scams and fraud, "an illegal call center building in a small city in Nigeria." For 3, "kids working in the informal economy in Belarus." His original includes a chart comparing attackers' wages with ordinary local wages. The excerpt doesn't reproduce it, so the actual figures aren't available here.
The headcount
Saxe's modeling produces a rough census of online crime:
About 720,000 people work in the online criminal economy.
About 90%, roughly 650,000, work in scam and fraud call-center operations. Some of them have been 4, meaning forced into the work.
About 76,000 are actual hackers, people who break into systems. Saxe says the "vast majority" of them work for governments.
That leaves a few thousand criminal hackers, and most of them work in ransomware.
The excerpt doesn't explain how these numbers were built. They also don't add up neatly. The government hackers appear to sit outside the 720,000 criminal total, and the exact count of criminal hackers isn't given. Treat them as order-of-magnitude estimates, not measurements. Saxe's piece also includes demographic estimates of these workforces, which Cowen's excerpt leaves out.
How this answers the question
Taken together, the argument works like a supply chain with a narrow point. Most of the damage that really hurts companies, such as breaking into networks and running ransomware, needs technical skill. People with that skill have good legal options or government jobs. So the criminal market for that skill is small, maybe a few thousand people worldwide. The far larger criminal workforce is low-skill labor running scripts on the phone. That labor can do a lot of harm one victim at a time, but it can't scale up sophisticated attacks.
Saxe's sharpest line is aimed at his own field. Security and AI-safety people, he says, sometimes imagine adversaries as their "5", an evil mirror image of themselves: "implicit MIT graduates." In reality, he argues, "in cybercrime we're dealing with a lower-skill demographic."
An analogy (ours, not Saxe's): think of the market for elite surgeons. If most of them work for national health systems, and the rest earn well in private practice, few will staff a black-market clinic. That clinic ends up run mostly by people with little training. It does harm, but it can't do complex operations at scale.
What's left open
The excerpt ends with a hint: "There are bottlenecks on the down side too…" Cowen doesn't say what those are. Presumably they are limits on how much criminals can actually do with access once they have it, but the text shown doesn't explain.
A few implications are worth flagging, and they are our inferences, not Saxe's claims:
Government hackers are a different kind of threat. If most skilled attackers work for states, the biggest risk may be espionage and sabotage, not theft. State hackers have different goals and aren't held back by what criminal gangs can afford to pay.
Cheaper skill could weaken this protection. If AI tools let low-skill workers do work that used to need experts, the hiring bottleneck Saxe describes could loosen. This is presumably why the argument matters to the AI-safety audience he addresses.
Fraud is the volume business. For individuals, the numbers suggest the most common threat is a scam call or message, not a clever hack. That fits the everyday advice: verify before you pay, and be suspicious of urgency.
The core claim is simple and can be tested. How dangerous criminals are depends partly on who they can afford to hire.
Remember
About 90% of the estimated 720,000 online criminals work in scam call centers, not hacking.
Most skilled hackers work for governments, which leaves only a few thousand criminal hackers.
Saxe argues the cost of hiring skilled people, not technology, is what limits how much damage cybercrime does.
